1. Field of the Invention
The present invention generally relates to an apparatus and method for treating toxic or hazardous substances, such as NO.sub.x and SO.sub.x or the like, contained in exhaust or flue gases of, for example, a thermal (electric) power plant, a garbage burning facility (namely, a refuse incinerating facility), a toxic substance treating facility and a car by using a streamer discharge plasma. More particularly, the present invention relates to a streamer discharge plasma treatment apparatus and method for decomposing and detoxifying nitrogen oxides (hereunder described as NO.sub.x) and sulfur oxides (hereunder described as SO.sub.x) contained in exhaust or flue gases of a thermal power plant and so forth. Further, the present invention is applied to the decomposition and detoxification of VOC (namely, Volatile Organic Compound) gases generated in a chemical factory or the like. Furthermore, the present invention relates in general to a pulse generator for use in the aforementioned exhaust gas treating apparatus and method, and more particularly, to a pulse generator which is useful as a power supply in the case that electrodes are placed in gases such as exhaust gases discharged from a thermal power plant or the like, that a (streamer discharge) plasma is then generated by delivering pulse power (or energy) to these electrodes (namely, by applying a pulse voltage across the electrodes) and that toxic substances are treated through electrical action.
2. Description of the Related Art
Hitherto, for instance, what is called an ammonia catalytic reduction method has been employed for decomposing NO.sub.x. Further, what is called a lime-gypsum method has been employed for decomposing SO.sub.x. Thus, what is called chemical processing or treatment methods (or processes) have been principal techniques for removing NO.sub.x and SO.sub.x, which are contained in exhaust or flue gasses, therefrom.
Meanwhile, in recent years, a streamer discharge plasma exhaust gas treatment method has come to be employed as such a technique. In an apparatus for treating toxic substances contained in exhaust gases by using a streamer discharge plasma, the streamer discharge plasma is generated in a (plasma) reactor chamber. The configurations of, for example, a conventional line-pair cylindrical reactor chamber A and another conventional line-pair flat-plate-like reactor chamber B are illustrated in FIGS. 15 and 16, respectively. Streamer discharge plasmas are generated in the reactor chambers A and B by applying (same) high voltages V.sub.0 across a line electrode 01 and a cylindrical electrode 02 of FIG. 15 and across a line electrode 06 and a plate electrode 08 of FIG. 16, respectively.
Electrons originated (or drawn) from a streamer discharge plasma are accelerated by an electric field, so that these electrons become high-energy ones. The high-energy electrons contained in the streamer discharge plasma decompose and detoxify toxic substances, such as NO.sub.x and SO.sub.x, which are contained in exhaust gases, by colliding with the toxic substances. For instance, in the case of decomposing NO, such high-energy electrons collide with NO and N.sub.2 to thereby induce the following reaction: NO+N.fwdarw.N.sub.2 +O. Thus, NO is decomposed.
Radical density contributing to the decomposition of NO is determined by energy cast or applied to the streamer discharge plasma. Moreover, the reaction rate of a reaction component of a reaction system is also determined (incidentally, note that the treatment rate of NO is physically determined when the radical density is determined).
In the conventional method or system, one high-capacity high-voltage power supply and one high-capacity reactor chamber are used so as to generate a streamer discharge plasma. Incidentally, in FIGS. 15 and 16, reference characters 04, 05, 010 and 011 designate electric current introduction lines.
However, in the case of a reactor, which has one high-voltage power supply and one reactor chamber as illustrated in FIGS. 15 or 16, constant energy is cast into the entire reactor chamber, regardless of the concentration of the toxic substance. Thus, there is caused an excessive waste of energy in a region, in which the concentration of the toxic substance is low, in an outlet of the reactor chamber. Consequently, the energy required for the treatment is increased. Namely, when the concentration of the toxic substance is lowered to a target value in a reactor chamber, energy of the amount, which is not less than the necessary amount of energy, is consumed in the region in which the concentration thereof is low.
Further, as compared with the conventional chemical processing or treatment methods, the conventional streamer discharge plasma exhaust gas treatment method has large merit in that the facility therefor is in-expensive and that a space required to install the facility is small. The conventional streamer discharge plasma exhaust gas treatment method has large demerit in that the energy consumption required for generating a streamer discharge plasma is about 10 Wh/Nm.sup.3 and is thus a little under two times that (namely, about 6 Wh/Nm.sup.3) required in the case of the conventional chemical processing or treatment method.
Meanwhile, a pulse generator for generating large voltage pulses has been used as a power supply for use in an apparatus for performing the streamer discharge plasma exhaust gas treatment method.
FIG. 17 is a diagram conceptually illustrating the configuration of a conventional pulse generator of the distributed constant (or parameter) type that uses coaxial cables. In this figure, reference characters 1.sub.-1 and 1.sub.-2 denote distributed constant (or parameter) lines (namely, transmission lines); 3 a high-voltage side wiring line (or wire); 4 a low-voltage side wiring line; V.sub.0 a D.C. charger; S.sub.1 a shortcircuit switch; V.sub.1-1 and V.sub.1-2 voltages generated in the direction of arrows corresponding to the distributed constant lines 1.sub.-1 and 1.sub.-2, respectively; Z a load; and V.sub.p a voltage applied to the load Z.
The distributed constant lines 1.sub.-1 and 1.sub.-2 are coaxial cables, whose characteristic impedances are Z.sub.1a and Z.sub.1b, respectively, and whose lengths are L. Further, each of the distributed constant lines 1.sub.-1 and 1.sub.-2 is composed of: a corresponding one of cores (or core lines) 1.sub.-1a and 1.sub.-2a ; and a corresponding one of outer conductors (made of shield braid or cladding materials or the like) 1.sub.-1b and 1.sub.-2b which surround the cores 1.sub.-1a and 1.sub.-2a through insulating materials (not shown), respectively. Incidentally, a folding-back point or portion is constituted only by the cores 1.sub.-1a and 1.sub.-2a that are not sheathed. These cores 1.sub.-1a and 1.sub.-2a are connected in series with each other. Further, an end (namely, an input-side portion) of these cores is connected to the D.C. charger V.sub.0 through the high-voltage side wiring line 3. On the other hand, the outer conductors 1.sub.-1b and 1.sub.-2b are connected to each other by a shortcircuit line 5.sub.-1 at the side of the shortcircuit switch s.sub.1 (namely, at the input-side terminal or end portion) and is thus shortcircuited. Moreover, the input-side terminal portion of the outer conductor 1.sub.-1b is connected to a grounding or earthing line serving as the low-voltage side wiring line 4. Furthermore, the input-side terminal portion of the outer conductor 1.sub.-2b is connected to the high-voltage side wiring line 3 through the shortcircuit switch S.sub.1.
In this case, the impedance of the D.C. charger V.sub.0 acting as a power supply is matched to that of the load Z.
Namely, Z=Z.sub.1a +Z.sub.1b.
Furthermore, in the case that the characteristic impedances of the distributed constant lines 1.sub.-1 and 1.sub.-2 are equal to each other, namely, in the case that the very same distributed constant lines 1.sub.-1 and 1.sub.-2 are used, the (voltage) propagation velocities of voltage signals in these distributed constant lines 1.sub.-1 and 1.sub.-2 are equal to each other. In the case where the dielectric constant of the insulating materials is .epsilon. and the magnetic permeability thereof is .mu., the voltage propagation velocity v is given by the following equation (1): ##EQU1##
In the case of such a pulse generator, the shortcircuit switch S.sub.1 is opened as an initial condition. Further, the high-voltage side wiring line 3 indicated by a thick line is charged by means of the D.C. charger V.sub.0 to the voltage V.sub.0. In this case, an output voltage V 1.sub.-1 of the distributed constant line 1.sub.1 -1 is V.sub.0, while an output voltage 1.sub.-2 of the distributed constant line 1.sub.-2 is -V.sub.0. Further, the voltage V.sub.p applied across the load Z is 0. The compositions of voltage waves respectively traveling in the distributed constant lines 1.sub.-1 and 1.sub.-2 are represented by sums of a progressive wave and a backward wave as indicated at moment t=0 in FIGS. 18(a) and 18(b).
Namely, FIGS. 18(a) and 18(b) illustrate the conditions of the voltage waves respectively traveling in the distributed constant lines 1.sub.-1 and 1.sub.-2 upon completion of operating (namely, turning on) the shortcircuit switch S.sub.1 at the moment t=0 after received. To put it more precisely, these figures corresponding to the moment t=0 represent the conditions of these voltage waves immediately before the shortcircuit switch S.sub.1 is turned on. At a moment t=L/2v, in the distributed constant line 1.sub.-2, the inversion of the polarity of the voltage wave occurs in a shortcircuit-switch-side part thereof, which is located at the side of the shortcircuit switch S.sub.1. In contrast, no variation in the voltage wave occurs in the distributed constant line 1.sub.-1, because both of the shortcircuit-switch-side terminal part and the load-side terminal part thereof, which are respectively located at the side of the shortcircuit switch S.sub.1 and the side of the load Z, are open ends (to be exact, both of these terminal or end parts thereof are regarded as open ends because exchanges of energy actually occur between the load and each of the distributed constant lines 1.sub.-1 and 1.sub.-2 but the exchanged energies cancel out).
Subsequently, at a moment t=L/v, the load-side terminal part of the distributed constant line 1.sub.-2, which is located at the side of the load Z, is put into a shortcircuited state, so that the voltage V.sub.p applied to the load Z becomes V.sub.0. This causes a variation in the voltage developed across the distributed constant line 1.sub.-1 1. As described above, the impedance of the charger V.sub.0 is matched to that of the load Z, the voltage waves travelling in the distributed constant lines 1.sub.-1 and 1.sub.-2 are not reflected by the end surfaces thereof but start propagating therefrom to the load Z. Further, a voltage generated in a time period between the moments t=L/v and t=3L/v is applied to and absorbed by the load, as illustrated in FIG. 18(c). As a result, the voltage, which has a peak value (or potential) of V.sub.0 and further has a pulse width of 2L/v, is supplied to the load Z, as illustrated in FIG. 18(c).
Thereafter, when the shortcircuit switch S.sub.1 is released or opened, the pulse generator is placed into the initial condition again. The process described hereinabove is performed repeatedly.
In the case that the peak (value of) voltage is raised by using the device illustrated in FIG. 17, a device configured as illustrated in FIG. 19 by stacking the devices, each of which is illustrated in FIG. 17, in such a way as to be independent of each other, suffices for such a purpose.
In FIG. 19, reference character V.sub.0 designates a D.C. charger; S.sub.1, S.sub.2, . . . , S.sub.N shortcircuit switches; 1.sub.-1, 1.sub.-2, 2.sub.-1, 2.sub.-2, . . . , N.sub.1, N.sub.-2 distributed constant lines; L the length of each of the distributed constant lines; Z a load; V.sub.p an output voltage applied to the load Z; 3 a high-voltage-side wiring line; 4 a low-voltage-side wiring line; and 5.sub.-1, 5.sub.-2, 5.sub.-N short-circuit lines. Here, note that pairs of outer conductors ((1.sub.2-2b, 2.sub.-1b) (2.sub.-2b, 3.sub.-1b), . . . , ((N-1).sub.-2b, N.sub.-1b)), each pair of which adjoin with each other as upper and lower stages, are connected with each other through connection lines 9.sub.-1, 9.sub.-2, . . . , 9.sub.-(N-1), serially, at the output-side terminal or end parts of the distributed constant lines 1.sub.-1, 1.sub.-2, 2.sub.-1, 2.sub.-2, . . . , N.sub.-1, N.sub.-2. On the other hand, the pairs of outer conductors ((1.sub.-2b, 2.sub.-1b), (2.sub.-2b, 3.sub.1b), . . . , ((N-1).sub.-2b, N.sub.-1b)) are not connected with each other at the input-side terminal parts of the distributed constant lines 1.sub.-1, 1.sub.-2, 2.sub.-1, 2.sub.-2, . . . , N.sub.-1, N.sub.-2, which are located at the sides of the shortcircuit switches S.sub.1, S.sub.2, . . . , S.sub.N. A circuit illustrated in FIG. 20 is an equivalent circuit of the circuit illustrated in FIG. 19. In FIG. 20, same reference numerals designate same components of the circuit of FIG. 19. Further, the (redundant) descriptions of such components are omitted herein.
Here, consider the case that all of the distributed constant lines 1.sub.-1, 1.sub.-2, 2.sub.-1, 2.sub.-2, . . . , N.sub.-1 and N.sub.-2 from the bottom stage (namely, the first stage) to the top stage (namely, the Nth stage) as viewed in this figure which have the same characteristic impedance and the same length, are used. Incidentally, in this case, it is assumed that the impedance of the load Z is matched to that of the power supply, namely, Z=Z.sub.1a +Z.sub.1b +Z.sub.2a +Z.sub.2b +. . . +Z.sub.Na +Z.sub.Nb.
In the initial condition, the shortcircuit switches S.sub.1, S.sub.2, . . . , S.sub.N are turned off. Moreover, the high-voltage side wiring line 3 indicated by the thick line is charged by means of the D.C. charger V.sub.0 to the voltage V.sub.0.
Upon completion of the charging, the shortcircuit switches S.sub.1, S.sub.2, . . . , S.sub.N are simultaneously turned on at the moment t =0. If the shortcircuit switches S.sub.1, S.sub.2, . . . , S.sub.N are completely simultaneously turned on, the voltage V.sub.p applied to the load Z at that time has the waveform as illustrated in FIG. 21. Therefore, a pulse, which has the peak voltage of NV.sub.0 and further has the pulse width of (2L/v), is supplied to the load Z.
FIG. 17 is a diagram conceptually illustrating the configuration of another conventional pulse generator of the distributed constant type that uses parallel flat plates. In this figure, same reference characters designate same portions of FIG. 6.
As shown in FIG. 22, the distributed constant lines 11.sub.-1 and 11.sub.-2 have flat plates 11.sub.-1a, 11.sub.-2a and 11.sub.-3, each of which has the same length L and further has the same width W. Further, the flat plate 11.sub.-3 is inserted between the flat plates 11.sub.-1a and 11.sub.-2a in such a manner that these three flat plates are parallel to one another. At that time, the flat plate 11.sub.-2 has the input-side terminal part connected to a terminal of the shortcircuit switch S.sub.1, which is opposite to the D.C. charger V.sub.0, and is grounded similarly as the flat plate 11.sub.-1a that has the input-side terminal part connected to the low-voltage wiring line 4 serving as a grounding line. The flat plate 11.sub.-3 has the input-side terminal part which is connected to a high-voltage-side wiring line 3. The flat plates 11.sub.-1a, 11.sub.-2a and 11.sub.-3 are insulated by dielectric insulating materials (or insulators), which are inserted between the adjacent flat plates (11.sub.-1a, 11.sub.-3) and between the adjoining flat plates (11.sub.-3, 11.sub.-2a), respectively, and have the same dielectric constant .epsilon., the same magnetic permeability .mu. and the same thickness D and further have the functions similar to those of a capacitor.
Thus, the distributed constant line 11.sub.-1, whose characteristic impedance is Z.sub.11a, is constituted by the flat plates 11.sub.-1a and 11.sub.-3, while the distributed constant line 11.sub.-2, whose characteristic impedance is Z.sub.11b, is constituted by the flat plates 11.sub.-1b and 11.sub.-3.
The capacitance C (of these distributed constant lines 11.sub.-1 and 11.sub.-2) in this case is given by the following equation: EQU C=2.epsilon.LW/D.
Thus, the capacitance C can be increased to a large value by changing the size (namely, the length L.times.the width W), the thickness D or the dielectric constant .epsilon..
Here, note that the impedance of the D.C. charger V.sub.0 is matched to that of the load Z, namely, Z=Z.sub.11a +Z.sub.11b. Further, the distributed constant lines 11.sub.31 1 and 11.sub.-2 have the same characteristic impedance. Namely, in the case of using the same distributed constant lines 11.sub.-1 and 11.sub.-2, the propagation velocities v of the voltage (wave) in these distributed lines are equal to each other.
Even in the case of the aforesaid pulse generator, which is thus configured by the parallel flat plates, the pulse voltage as illustrated in FIG. 18(c) is supplied to the load Z by the action similar to that in the case of the pulse generator configured by the coaxial cable of FIG. 5.
In the case that the peak voltage is raised by using the circuit illustrated in FIG. 22, a device configured as illustrated in FIG. 23 by stacking the circuits or units, each of which is illustrated in FIG. 22, in such a way as to be independent of each other, suffices for such a purpose. This device corresponds to the device illustrated in FIG. 19. Thus, in FIG. 23, same reference numerals designate same portions of FIG. 19. Further, the redundant description of such portions is omitted.
The pulse generator of FIG. 23 is obtained by stacking up N (stages) of the pulse generators, each of which has the structure shown in FIG. 22 and is used as a component unit. Namely, this device has the N stages of the distributed constant lines 11.sub.-1, 11.sub.-2, 12.sub.-1, 12.sub.-2, . . . , NN.sub.-1, NN.sub.-2, which are constituted by the parallel flat plates. Incidentally, in each pair (or stage) of the upper and lower stages, a single flat plate serves as both of the top flat plate of the lower stage and the bottom flat plate of the upper stage.
Even in the case of such a pulse generator, the pulse voltage illustrated in FIG. 21 is generated by simultaneously turning on the shortcircuit switches S.sub.1, S.sub.2, . . . , S.sub.N upon completion of predetermined preparation, similarly as in the case of the pulse generator of FIG. 19.
However, in the case of generating pulses be means of the pulse generators configured as illustrated in FIGS. 19 and 23, it is necessary to simultaneously turn on the shortcircuit switches S.sub.1, S.sub.2, . . . , S.sub.N, though the turning-on operation of each of these switches is not performed in exact timing (namely, in perfect synchronization) with those of the other switches and thus there is observed a phenomenon in which the output voltage V.sub.p drops. This is because of the facts that the pulse width is of the order of nanoseconds (ns) and is thus extremely short, that therefore, the influence of delay in the application of a trigger voltage or in discharge due to the difference among the wiring impedances respectively corresponding to the shortcircuit switches S.sub.1, S.sub.2, . . . , S.sub.N is enhanced and that it is difficult to time the operations of turning on the shortcircuit switches (namely, it is difficult to simultaneously turn on the N switches).
If there is caused the delay in the application of the trigger voltage, output voltages V.sub.p of three circuits have waveforms, for example, A, B and C as illustrated in FIG. 24. Thus, a synthesis pulse (A+B+C) synthesized from these three pulses A, B and C has a waveform as shown in this figure. Consequently, a pulse, which has an ideal waveform and has a pulse width of 100 ns and a peak voltage 3V.sub.0, is not obtained.
The present invention is accomplished in view of the aforementioned various problems of the conventional apparatuses and methods.